EASA Moves AI Rulemaking Toward Aviation-Specific Requirements

The European Union Aviation Safety Agency is advancing its work on artificial intelligence regulations, aiming to adapt its general AI framework to specific areas of aviation. The agency's rulemaking task RMT.0742 covers requirements for AI across regulated activities, including aerodromes. An update on the task is scheduled during EASA AI Days in Cologne.

EASA published its first proposed AI rules last November. The proposal sets baseline requirements for AI assurance, human factors, and ethics for systems that assist or cooperate with humans. Comments on the proposal closed in March. A second proposal will apply the framework to individual aviation domains.

In June, EASA released an updated concept paper that expands scope to reinforcement learning, symbolic AI, and Level 3 advanced automation. The agency noted these applications enable operations where the human is remotely present or absent.

The conference is also featuring 10 AI systems selected from 50 submissions for its first AI Olympics. Finalists include Kairós Aeroespacial's FOR-DEC AI, a cockpit tool for organizing operational information, comparing options, and identifying uncertainty in decision-making. Other finalists analyze aviation safety reports and address certification.